Assimilating high temporal frequency SPOT data to describe canopy functioning: the ADAM project

The objective of this study is to develop a methodology for estimating canopy state variables in time and space using assimilation of high temporal frequency SPOT data into vegetation process models. This work is part of the ADAM (Assimilation of spatial Data within Agronomic Models) project, a Franco Romanian partnership which experiment provided a complete scientific database of remote sensing and ground data products. The canopy functioning model chosen is STICS (Simulateur mulTIdisciplinarie pour les Cultures Standard) developed to simulate a wide range of crops including wheat. The ADAM experiment will be briefly described along with the approach used to assimilate SPOT observations. A discussion is drawn on the splitting between the potential sources of variation of the parameters: those associated to the cultivar, those that vary between fields such as cultural practices, and those that very within fields such as soil characteristics. The assimilation is based on a variational method for which we have computed the adjoint model of STICS. This tool will enable to perform a sensitivity analysis, with regard to the observations and to the input parameters, to finally investigate the effects of high temporal frequency.